This project extends the Center model of PFC-MTL interactions in context-guided associative memory to the domain of spatial navigation in rats. As in the other projects, we will test the hypothesis that the PFC develops generalized representations of context abstracted from more specific MTL codes, and that context dependent associations coded by hippocampal neurons depend upon specific PFC circuits. This project will employ a testing protocol to assess context guided memory that is formally identical to those of other projects. However, here we will investigate context-guided learning and selection of spatial routes, thereby linking well investigated MTL mechanisms of spatial navigation and path integration with PFC mechanisms that support abstract rule learning and reward expectancies. The project design corresponds closely with Project 4, which uses contextual cues to guide object reward associations, and this close correspondence will allow us to directly compare mechanisms of spatial and nonspatial context-guided associations. The specific aims combine high density recording and temporary inactivation in targeted PFC and MTL regions in behaving rats to determine the functional and physiological coding contributions of pairs of interacting frontal and temporal networks. In each aim, rats will be trained to select different routes depending on the environmental context. Aim 1 will analyze multiple single units and local field potentials (LFP) in pairs of tetrode bundles placed in PFC and MTL targets. Aim 2 will use either bilateral or crossed ipsilateral temporary inactivation of the same brain regions to investigate their causal role in context selective routes. Aim 3 will explore how coding features in one region depend upon the other by combining local circuit inactivation and recording. Results will help interpret the mechanisms of brain impairment and activity described in humans and monkeys in Projects 1-4. The Analytic Core will compare the data from Projects 3-5 to quantify the similarities and differences in neural coding across species and types of contextual associations, and the project will help constrain and be guided by the computational modeling in Project 6.